The Results of my Preamp Trials
Posted by: Alex S. on 23 October 2001
The 3 preamp combinations tested were 32.5/SC; DNM 3B Primus; Dynavector L100. Apart from the Naim PSU all three are now superceded by newer models. Approximate cost to buy each new would be, 32.5/SC, £2500; DNM, £1800; DV, £2000.
Only the preamp and connecting cables were variables, the rest of the system remained the same: CDS2/CDPS; 250; B&W N805s. The 32.5 was connected with Naim cabling, the DNM with dnm cables and the DV with cables made up by the Chord company.
All equipment was warmed up/run in. Normal, blind and partially sighted tests were carried out. I had two friends to help me listen (one an esteemed technical guru from this forum, the other a maker of drums) and also a guy from Dnm. It should be said that the gain on the Dnm may not be set up perfectly for the CDS2.
Vinyl was not tested since the DV is line only and the Dnm phono stage had never been used and was therefore regarded as unfairly disadvantaged.
So that I don't go on and on I have devised a numerical system for the results: marks out of ten for a variety of attributes. If you consider such a system totally meaningless, fine.
Nothing I say here has anything whatsoever to do with my annoyance that Blzebub was banned from the Forum, but it is fair to warn those who don't know that I never got along with the 82 and that I think the 52 to be a very fine preamp indeed.
The Results:
32.5/SC:
Pace 7, Rythm 8, Timing 7.5, Sweetness/Musicality 6, Dynamics 8, Clarity 6, Quietness 5, Round Earth 7.5, Bass Depth 6, Bass Control 6.
Total Marks (out of 100): 67
Dnm 3B Primus:
Pace 6, Rythm 7, Timing 8.5, Sweetness/Musicality 7, Dynamics 8, Clarity 8.5, Quietness 7, Round Earth 8, Bass Depth 8, Bass Control 6.5.
Total Marks (out of 100): 74.5
DV L100:
Pace 8, Rythm 9, Timing 10, Sweetness/Musicality 8, Dynamics 9, Clarity 9, Quietness 9, Round Earth 8, Bass Depth 7.5, Bass Control 8.
Total Marks (out of 100): 85.5
A few thoughts:
1. There was no compatability problem with the 250, running nice and cool all the while.
2. The Dnm would probably be better with Dnm or Crimson power amps and could also be better set up. Equally, the DV would probably sound better with DV power.
3. Naim amplifiers can do round earth stuff its the backs to the wall speakers that can't.
4. To my mind the DV would give the NAC 52 a very stern test at a fraction of the cost.
5. The DV L300 which replaces the L100 could be superb.
And here's (most of) who we listened to: Ricky Lee Jones, Bill Ray Vaughan, Ozric Tentacles, Moloko, Massive Attack, Porcupine Tree, Faithless, Dreadzone, Clash, Spock's Beard, Terry Bozzio, Steve Stevens, Beethoven, Mahler, Berlioz, Shostakovich.
[This message was edited by Alex S. on TUESDAY 23 October 2001 at 09:51.]
The tables didn't work. Should be a bit clearer now.
Andrew,
I've already done extensive tests of 32.5 with HC, an un-named PSU and the Supercap. A 32.5/HC although more logical would struggle to make 50 in these tests.
Alex
A dealer I know prefers it vastly to the 52..
Good review of the other stuff too...
( now what more music can you recommend to me :-) )
Cheers
Goose
quote:I wonder if that's the same dealer I know..
A dealer I know prefers it vastly to the 52..
Alex
ps I am working on some sort of sensible music response.
Otherwise interesting. Try it with a different power amp and I bet the results change again - it's a tricky business this hifi lark!
Also interesting that you found the pace of the DNM to come last - to my ear it was the fastest preamp I'd ever heard (but again, I've only heard it in the context of my own system, which has perhaps a faster poweramp...)
John
I know you have chosen a Dnm and I applaud your choice. In my system it happens to under-perform probably because of the set up, the 250 or the N805s. Also my preamp is un-Mana'd. I would say that Andy and I thought it could become tiring, the dynamics being a little forced. Thinking about it more, I guess it deserves 7 for speed. Also, it is very much last year's model if that makes a difference. A Dnm is quite a departure from the Naim sound I am used to for sure.
I found the 82 to do all the Naim stuff very well I just didn't like its presentation - uncouth and aggressive and no great control in the bass.
I found the 32.5 about the same as 82 for pr&t, but warmer. It doesn't have many 'hi-fi' virtues and is noisy. Image and soundstage are surprisingly good though.
JW
I've heard a 52 five times but never in my own system. I think the DV L100 is virtually its equal but that is instinct and not based on any tests. Instinct also tells me that the L300 could take top honours.
Brian
The DV slots beautifully into a Naim system (one reason its been chosen). It does the Naim stuff really well and times better than anything I've heard. It is incredibly quiet and separates notes as well as a 52. AFAIR the 52 beats it in terms of the leading edge of notes and the handling of transients. The 52 defines instruments better probably but the DV runs it very close.
IMO the DV outperforms an 82/SC in every department and has wonderful bass control if not bags of bass.
The 250 runs barely warm at high volume (it'll probably blow up tomorrow and you can all laugh).
All I can really say is that DV warrants a serious listen. It suits me, but hi-fi is a subjective business. I'm sure many of you think I'm talking rubbish and that the 82 is far superior.
Alex
It's fair to say the Dynavector pre-amp has turned all of my own ideas about designing world-class preamps on its head. Most of those ideas come from my experience and indoctrination of the laws of audio according to Naim.
DV have considerately scratched the markings from all of the important devices inside the pre, probably because there's nothing unique about the components used. Now are you sitting down, 'cos this bit might shock you?
I reckon the DV uses op-amps!
It uses IC based electronic input switching, and I'd hazard a bloody good guess that the active devices are op-amps. There's little discrete circuitry and the pre talks about it's low impedance output, capable of driving long cables. This comes from a 5 legged TO220 device which I reckon, horror of horrors, is a power op-amp.
I even looked under the PCB expecting to see some SM stuff, but it wasn't there!
It has an internal, but shielded toroid, and extensive regulation on-board, smoothing caps are modest (i.e. small by Naim standards).
It sounds bloody marvellous though, and has set me thinking about designing something similar, since there are some fabulous op-amps for audio on the market at present, and let's face it they're easy to work with.
As for bandwidth limiting, I didn't get chance to examine in that kind of detail, but there definitely isn't anything of the complexity of the time-aligned filters. Any roll-offs will be similar to those of early Naim amps, just simple RC circuits.
Since Alex is no longer using Naim amps either I don't think the bandwidth issue is a problem.
Andy.
I'm afraid my heresy continues: the 250 now seems a limiting factor. New power amps are under test and consideration. And I do fear a final topic called the results of my power amp trials to come. (And finally the speakers, if I'm not banned by then).
I promise to all that the CDS2 is going nowhere!
Alex
quote:
Originally posted by Alex S.:
I'm afraid my heresy continues: the 250 now seems a limiting factor. New power amps are under test and consideration.
As per my e-mail, I still feel that much of your problem with the Naim sound is caused by ferrous stuff near the system (Mana, Sound Org, LP12 etc).
Sorry!
Martin
quote:Can the presence of iron make that much difference? And why does it foul up Naim's equipment to the exclusion of others'? In any case the Base is fairly non-ferrous and the Mana, SO and LP12 are at least a metre away from any Naim stuff.
I still feel that much of your problem with the Naim sound is caused by ferrous stuff near the system (Mana, Sound Org, LP12 etc).
quote:As you know I know sod all technically but what Mr T says seems borne out by the DV facts. The immediate, and very satisfying first impression is just how quiet the DV is. Obviously, this is not an end in itself but I'm sure its a means to a fine end.
Ultra low noise.
High PS noise rejection.
V high i/p, V Low o/p impedance.
Ultra high slew rate.
Ultra low distortion.
Would an op amp design be expected to show exceptional timing? Because the DV certainly does.
Alex
ps Martin, best not say this sort of thing in Mana land if you wish to remain un-insulted.
quote:
Originally posted by Alex S.:
Can the presence of iron make that much difference?
Well, yes.
When I tried out the Base/Starbase racks we set the system up (PS's & LP12 on one rack, unpowered items on a second, 135s on a third) and put on the first track. Awful.
All the harshness that you discuss was there, together with a disjointed presentation.
Took the LP12 away (which was plugged into the 52, but not into the mains) and the improvement was remarkeable.
The only problem is if you find this was the "problem" with your 82, and regret selling it. I have heard exactly this effect moving an 82 based system (Juan Zenuff's) from Sound Org to wooden racks.
As I said, sorry.
Martin
I've heard the LP12 messes things up before but surely moving it backwards and forwards each time I change the source is going to mess it up something rotten.
I've done a lot of messing about over the last months and I do believe that its the boxes themselves and the room acoustics which determine a basic sound. Changing cables, supports, adding cones, putting everything in a different order on said support, cryogenic freezing of speaker cable whilst annealing the interconnect all makes a difference, sometimes quite a large one, but it does not change an inherent signiature tune.
Fraim is here and suddenly supports are vital but don't forget that Naim spent at least 20 years saying that you could dump all your Naim boxes on the shagpile and they'd still sound great. They were right. My second system is on vinyl flooring, carpet and Ikea cheapo. It sounds good. The only truly vital stand is the one for the speakers.
Alex
quote:
Originally posted by Paul Duerden:
When I first heard a Naim system it was on highly ferous sound Organisation stands. I was very impressed by the sound of the system and bought a 62 110 which eventualy turned into a 62 Hi Cap 250. In those days Naim were perfectly happy with ferous supports as were their dealers .Is the newer Naim kit more sensitive to ferous objects than the old stuff was? I mean if the ferous support so terminaly buggers up the sound of a Naim amp, those who bought them years ago must be deaf as posts, it also asks the question of how Naim managed to expand as a company.
I agree that this is a major issue, and it's something I struggle to understand. However, audio memory is much more fickle, rose-tinted even, than we give credit for. I bet if you took a good system on Hutter/Mana/Quad, etc (you choose) and put it back on Sound Org you'd be in for a real shock.
I wonder if part of the issue is the proliferation of RF pollution. Think of the 10,000 or so mobile phone masts in the UK (about to be quadrupled for 3G). Think of all the stuff in your house with embedded microprocessors (microwave, central heating controller, clock/radios etc, etc). This RF then causes designers to put mains filters into their kit. You may have a dozen or more items with Lingo-style mains filters in them permanently plugged into the mains (all the above microprocessor stuff, and PCs, phone chargers, TV, remote control table lamps, etc, etc, etc).
Do ferrous objects somehow making the kit more susceptible? I have seen reports that earthing of metal racks leads to great benefits for some (haven't tried it myself), which would certainly support this argument.
Bear in mind that I'm running active, so have a lot of cabling trailing around the place.
Also, there was a subtle change to the Naim sound around the time the "new" case style came in. The NAC72 was less warm than the 32.5/42.5/62, and the new transformers introduced a different sound, more dynamic and tight, but less warm. My memories of the move from old-style 32.5/hi/140 to new-style 72/hi/250 in one go are of struggling to calm down an 'edge'. Unrealised ferrous problems, perhaps? A friend still owns this kit, and it sounds great in his environment.
I presume your original experience was with a vinyl source? These still seem softer to me than CD.
For all the brilliant performance I've had from my system over the years, there have always been some tracks that just seem too in-your-face. It seems I can now happily play a wider range of music. Even the tracks that worked well with a bit of favouritism work even better when reproduced in an even-handed manner. This is not to say my system is now fully sorted in the current house, but the new racks have moved me another large step forwards.
quote:
Also what about speakers, anyone using ATCs DBLs etc have magnets that you just would not believe. I have seen a 12inch Atc driver and the magnet is gigantic. This must surely affect equipment a sight more than a ferous stand.
So far, I have not detected any problems when the metal is kept 2-3 feet from the system. I suspect that if the DBLs were that close to the kit it wouldn't last long.
quote:
Then there is the problem that if your stand is against a wall there is a strong possibility that the equipment is close to ferous objects. Near a radiator would be a complete no no.
Yup.
Audio Counsell also tell us that placing the gear near to mains sockets or blocks is bad. Difficult for me to get away from this, at the moment.
quote:
How do ferous objects affect my own Densen amp? I have played about with this when I briefly had the opportunity to try a Quaraspire rack, With neither my Micromega CD player nor my densen amp , could I detect a difference when a large metal weight was put on top of each one respectively. Of course I dont know what is in the wall in front of which the Quadraspire had been placed. This is obviously an unknown variable.
Are the cases of the Densen non-ferrous?
However, I had found some benefit removing ferrous items from near the kit even when using ferrous stands.
I'm told Linn were banging on about this issue nearly 10 years ago, although I wouldn't implicitly trust anything they say, given what I've heard at shows recently.
quote:
You will forgive me for not having the Linn in that test. Duty does not extend to taking large lead weights anywhere near a MC cartridge.
At least lead is non-ferrous!
cheers, Martin
quote:For your benefit I just did. Its shockingly similar.
I bet if you took a good system on Hutter/Mana/Quad, etc (you choose) and put it back on Sound Org you'd be in for a real shock.
quote:May I suggest that maybe, just maybe, you have answered your own question here.
Also, there was a subtle change to the Naim sound around the time the "new" case style came in. The NAC72 was less warm than the 32.5/42.5/62, and the new transformers introduced a different sound, more dynamic and tight, but less warm. My memories of the move from old-style 32.5/hi/140 to new-style 72/hi/250 in one go are of struggling to calm down an 'edge'.
The new transformer has a lot to answer for IMO. I don't remember sudden meteorite showers and Mana falling from Heaven when the new Naim sound started to appear.
Alex
quote:
Originally posted by Alex S.:
For your benefit I just did. Its shockingly similar.
Selling the Mana then?
quote:
May I suggest that maybe, just maybe, you have answered your own question here.The new transformer has a lot to answer for IMO.
You may be right, but the new stuff is capable of a better ultimate performance. It's just that the greater bandwidth and dynamics can take you into room problems and is even more revealing of source problems. Makes source first even more important to control the extremes.
cheers, Martin
Your knowledge is impressive.
I will once again reiterate ... are you a thrusting dynamic recent graduate or do you,in fact,work as a technical director of a well-know audio concern?
In any event, keep up the good postings I am enjoying them.
AQD
quote:
Originally posted by Mr Tibbs.:
Think of the op-amp as a 'black box' or 'building block', similar in design to any power amp. It requires very few additional components to do all manner of duties in a pre-amp or active crossover.
Early op-amp designs sounded pretty awful it has to be said, but current SOTA op-amps seem to neither add or take away from the signal passing through.
I'm not saying discrete technology can't do better, Naim have used discrete design in the cds2, but it is probably very similar to recent op-amp design.
The successor to the 52 will probably use this technology, with a cheaper stablemate using op-amps. The trouble is, a dual rail type supply is needed so Hi-caps etc. Become redundant.
Op-amp based designs can be very compact and as a result may not be subject to the problems discrete designs seem to suffer, in terms of vibration or electromagnetic effects.
Mr Tibbs.
Time for some Stella Artois I thin
quote:
Op-amp based designs can be very compact and as a result may not be subject to the problems discrete designs seem to suffer, in terms of vibration or electromagnetic effects.
Naim always used to use discrete circuitry because it was supposedly less susceptible to vibration than IC's. Still to bloody sensitive though!
The biggest improvement to modern op-amps' performance, when used in audio, are models with much greater open loop bandwidth. Ideally one needs to cover the audio band before open-loop rolloff. The difference is clear when comparing otherwise similar amps.
Look at the AD825, or even the BB amps used by Naim - the AD825 is a fabulous sounding op-amp. There may be even better devices now, it's just you often cannot do parametric searches on the parameters of interest.
Therefore it takes bloody ages to find good ones.
As for discrete op-amps (a la CDSII) I think these are even used in the NAP150 (I assume they're the black resin-potted rectangular devices, near the input. I assume they're replacing the differential input pair?
Andy.
On the subject of op-amps I have a bit of an aversion to them. The reason is exactly what Mr T mentioned earlier, that they are similar in design to power amps. I agree that the more feedback you can lash around them the better they track the input signal, provided they remain stable, but there are architectural limitations that I think are unecessary for an audio circuit.
Firstly, they usually use push-pull output stages to reduce power dissipation. This is a problematic topology wrt distortion. Secondly, all the gain stages have common power rails - local isolation is not practical due to the physical size of reactive components when implemented in silicon. They also tend to use a lot of transistors and other junctions in place of resistors because of the difficulty of implementing pure resistances in silicon. Junctions are non-linear. A number of compromises are made to serve size and silicon fabrication constraints and power dissipation. In discrete circuits these limitations go away.
Other than size and convenience and, to some extent, consistency of specification across batches I would have thought a higher performance circuit, for audio, can be made discretely. So why use an op-amp?
quote:Never knew Lancaster was that rough, or is it just a local reaction to your 'early music'?
in the case of my terrace what looks like iron bars in them (the walls)
You don't think I reallymoved everything onto the SO rack do you?
Listening yesterday to Pog's rig I sort of decided that the presence of ferrous material (enough of it to sink a small ship) may not be quite so bad after all!
Alex
quote:
Originally posted by Mr Tibbs.:
The 'designed for audio' opamps available now have little or no 'sonic signature' so they can be used liberally without any penalty in sound quality terms. They have many benifits apart from obvious ease of use,
Ultra low noise.
High PS noise rejection.
V high i/p, V Low o/p impedance.
Ultra high slew rate.
Ultra low distortion.
The list is endless. It is very difficult to come up with a discrete design capable of bettering the performance of todays not so humble opamp.
quote:
Date: 29-Mar-98 07:11
Author: julian vereker
Subject: op amps
Op amps suitable for audio didn't exist when I designed our power amp circuit and even today one can probably do better with discrete transistors.julian
It's all a question of time and effort though, and a well chosen modern op-amp can be a superb solution for a quick and simple design.
The CDSII, NAP500 (+NAP150?) use Naim's 'discrete op-amp' I suggest it's just that - a discrete circuit designed for a specific circumstance.
The stability margins required for more general solutions are what limits most op-amp designs. Reduce the range of application and it get's easier - op-amps are a jack of all trades, master of some.
JV's statement still holds true, and probably will for ever.
Andy.
quote:
Originally posted by Mr Tibbs.:
Just thought I'd let you into a wee secret, the design of the NAP 500 'probably' has ripped off a few ideas from recent op-amp technology.
This statement reads to me as "Naim used discrete components, presumably because this was better than an off-the-shelf op-amp".
If this isn't what you meant you'll need to rephrase.
cheers, Martin
Regards,
Bah